[Skip to Content]
[Skip to Content Landing]
Views 235
Citations 0
Comment & Response
May 20, 2019

Laterality and Stimulation Bias in Meta-analysis of Placebo Responses

Author Affiliations
  • 1Spaulding Rehabilitation Hospital, Charlestown, Massachusetts
JAMA Neurol. 2019;76(7):869-870. doi:10.1001/jamaneurol.2019.1229

To the Editor Zunhammer et al1 claim that the effects of placebo are small in terms of changes in bottom-up nociceptive processing based on an analysis of participant-level neuroimage contrast maps from 603 healthy participants in 20 placebo pain imaging studies. We commend the authors and the contributing groups for sharing their findings. However, we have found a bias in how they analyzed this source. Briefly, they computed the dot product of individual placebo contrast images with a predefined weighted set of brain regions called the neurological pain signature (NPS). From this computation, they calculated the effect size (Hedges g) within each of the 20 placebo studies. Their results indicated a large effect for noxious stimulation (g = 2.30) on the NPS outcome and a moderate effect of placebo on pain ratings (g = −0.66), but only weak evidence that placebo altered the NPS (g = −0.08).